This invention relates to gene interaction network construction. More specifically, the invention relates to combining information from two information modalities for gene interaction network construction, the former being clustering of gene expression data, and the latter being derived from morphological similarity of RNA Interference (RNAi) imaging data from the same genes.
Research in the area of biotechnology is constantly growing and evolving. Construction of a gene interaction network is sought after to provide knowledge and understanding of biological processes and diseases, and in the future may help with the identification of therapeutic compounds. Building gene interaction networks and predicting functions of genes poses a significant challenge to scientists. Model organisms are known to contain a large number of genes whose functions are unknown. Pathways, such as those for apoptosis and mitochondrial biogenesis, have only been partially identified. The rate of gene function determination significantly lags behind gene sequencing, thereby causing a backlog of genetic sequences in search of a function.
A number of approaches have been used to reveal genetic pathways, ranging from traditional low-throughput genetic screens based on phenotypes to microarray experiments that reveal gene expression similarity under various experimental conditions. In particular, visual assays using microscopy represents a medium for phenotypically assessing gene function that is complementary to microarrays. With the increasing use of RNAi through modern three dimensional restorations or confocal microscopy and high throughput imaging platforms, large amounts of image based data has become available. Microscopic imaging is a powerful tool that allows functions and aspects of cells to be studied via morphological changes, such as changes in size and shape of cells, number of subcellular organelles and structures, or the redistribution of specific proteins within the cell.